Aalborg Universitet
Lithium-Ion Battery Operation, Degradation, and Aging Mechanism in Electric Vehicles:
An Overview
Guo, Jia; Li, Yaqi; Pedersen, Kjeld; Stroe, Daniel-Ioan
Published in:
Energies
DOI (link to publication from Publisher):
10.3390/en14175220
Creative Commons License CC BY 4.0
Publication date:
2021
Document Version
Publisher's PDF, also known as Version of record Link to publication from Aalborg University
Citation for published version (APA):
Guo, J., Li, Y., Pedersen, K., & Stroe, D-I. (2021). Lithium-Ion Battery Operation, Degradation, and Aging Mechanism in Electric Vehicles: An Overview. Energies, 14(17), [5220]. https://doi.org/10.3390/en14175220
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
- Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
- You may not further distribute the material or use it for any profit-making activity or commercial gain - You may freely distribute the URL identifying the publication in the public portal -
Take down policy
If you believe that this document breaches copyright please contact us at vbn@aub.aau.dk providing details, and we will remove access to the work immediately and investigate your claim.
Review
Lithium-Ion Battery Operation, Degradation, and Aging Mechanism in Electric Vehicles: An Overview
Jia Guo1,2 , Yaqi Li1,2, Kjeld Pedersen2and Daniel-Ioan Stroe1,*
Citation: Guo, J.; Li, Y.; Pedersen, K.;
Stroe, D.-I. Lithium-Ion Battery Operation, Degradation, and Aging Mechanism in Electric Vehicles: An Overview.Energies2021,14, 5220.
https://doi.org/10.3390/en14175220
Academic Editor: Tek Tjing Lie
Received: 28 July 2021 Accepted: 18 August 2021 Published: 24 August 2021
Publisher’s Note:MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1 Department of Energy Technology, Aalborg University, 9220 Aalborg, Denmark; jgu@energy.aau.dk (J.G.);
yaqili@mp.aau.dk (Y.L.)
2 Department of Material Production, Aalborg University, 9220 Aalborg, Denmark; kp@mp.aau.dk
* Correspondence: dis@energy.aau.dk; Tel.: +45-99403327
Abstract:Understanding the aging mechanism for lithium-ion batteries (LiBs) is crucial for optimiz- ing the battery operation in real-life applications. This article gives a systematic description of the LiBs aging in real-life electric vehicle (EV) applications. First, the characteristics of the common EVs and the lithium-ion chemistries used in these applications are described. The battery operation in EVs is then classified into three modes: charging, standby, and driving, which are subsequently described.
Finally, the aging behavior of LiBs in the actual charging, standby, and driving modes are reviewed, and the influence of different working conditions are considered. The degradation mechanisms of cathode, electrolyte, and anode during those processes are also discussed. Thus, a systematic analysis of the aging mechanisms of LiBs in real-life EV applications is achieved, providing practical guidance, methods to prolong the battery life for users, battery designers, vehicle manufacturers, and material recovery companies.
Keywords:lithium-ion battery; electric vehicles; aging mechanism; battery degradation
1. Introduction
Lithium-ion batteries (LiBs) with high energy density are receiving increasing atten- tion because of their environmental friendliness and are widely used in electric vehicles (EVs) worldwide [1]. Battery degradation problems, such as capacity fading and internal resistance increasing, inevitably occur with time and use. These cause great trouble to users and manufacturers [2]. A clear understanding of how batteries age in EVs is urgently needed to: (i) optimize the battery materials, (ii) improve battery cell production, and (iii) guide the design of automotive battery systems.
At present, scientists from different fields have researched, from different perspectives, the aging of LiBs. Some scientists specifically discussed the impacts of environmental and op- erational factors on battery degradation [3], while others studied the battery aging mechanism through the post-mortem analysis of the internal components of the battery cell [4]. However, a close connection between the battery operation and degradation in EV applications and the corresponding aging mechanism has not yet been established. Thus, a review is necessary in order to systematically and comprehensively describe the aging of LiBs in EVs.
Many reviews on battery aging have been published presenting the battery degra- dation and aging mechanisms. The main contents of these reviews are summarized in Table1. These reviews are mostly based on analyzing laboratory accelerated aging test results, which are mainly obtained using constant charging/discharging current and are significantly different from the battery operation in EVs. Besides, most of them lack the con- nection with the battery operation scenarios, and focus only on the degradation behavior of the battery itself; in reality, the influential factors on battery charging, discharging and standby are different, and aging should be described independently based on the operation status. Moreover, the battery chemistries reviewed in these works mainly involved stable
Energies2021,14, 5220. https://doi.org/10.3390/en14175220 https://www.mdpi.com/journal/energies
Energies2021,14, 5220 2 of 22
LiCoO2and LiFePO4, which are more stable and mature and are not considered to be state- of-the-art technology for EVs. Therefore, the aging mechanisms of widely EV-used Ni-rich battery chemistries (LiNi1−xMxO2, M = Co, Mn and Al. (NMC) and (NCA)) need further study. In order to address this gap, in this paper, we review the NMC and NCA battery operation, degradation, and the corresponding aging mechanisms in real-life-EV use.
Table 1.An overview of the published literature related to battery aging.
References Topic
Chemistries Operation Degradation Aging Mechanism
Han et al., 2019 [5] √ √
LMO, LCO, LFP, NMC
Tian et al., 2020 [6] √
LMO, LFP, NMC
Mocera et al., 2020 [7] √
LFP
Woody et al., 2020 [8] √ √
LCO, LMO, LFP, NCA, NMC
Vetter et al., 2005 [9] √
LCO, LMO, NMC
Broussely et al., 2005 [10] √
LCO, NMC
Barre et al., 2013 [11] √ √
LCO
Birkl et al., 2017 [12] √
LCO
Palacin et al., 2018 [13] √
LMO, LCO, NMC, NCA
Xiong et al., 2020 [14] √ √
LFP, NCA, NMC
Teichert et al., 2020 [15] √
NMC
Alipour et al., 2020 [16] √ √
LCO, LFP, NMC, NCA
Chen et al., 2021 [17] √ √
LCO
Yang et al., 2021 [18] √ √
NMC, NCA
In this article, we analyzed the applications of LiBs in current EVs, and divided the battery operation scenario into three modes: charging, standby, and driving. The influence on EV battery degradation from the corresponding factors for these modes is studied, respectively. Finally, the relationship between the battery operation mode and the aging mechanisms of battery cell components (i.e., anode, cathode, and electrolyte) is established in order to clearly describe LiBs aging in real EV use.
The remainder of this paper is structured as follows: in Section2, we will introduce the LiBs chemistries used in EV applications. Section3will describe in detail the battery operation modes in EV application. Section4will present the LiBs degradation in EV applications. Section5will discuss the aging mechanisms of LiBs, which are caused by the EV operation, and Section6provides concluding remarks.
2. LiBs in EV Applications 2.1. Electric Vehicle Applications
Depending on the primary source of energy, EVs can be classified mainly into the following three types: Battery Electric Vehicles (BEV), Hybrid Electric Vehicles (HEV) and Plug-In Hybrid Electric Vehicles (PHEV). The energy flow in these three types of EVs is presented in Figure1.
Energies 2021, 14, x FOR PEER REVIEW 3 of 23
2. LiBs in EV Applications 2.1. Electric Vehicle Applications
Depending on the primary source of energy, EVs can be classified mainly into the following three types: Battery Electric Vehicles (BEV), Hybrid Electric Vehicles (HEV) and Plug-In Hybrid Electric Vehicles (PHEV). The energy flow in these three types of EVs is presented in Figure 1.
Figure 1. The schematic representation of the three types of EVs; from left to right: Battery Electric Vehicles (BEV), Hybrid Electric Vehicles (HEV), and Plug-In Hybrid Electric Vehicles (PHEV).
BEVs are completely powered by electrical energy from batteries. The external energy is supplied only by plugging the BEV into the electricity grid for charging, and there is no on-board electricity generation. Depending on the EV manufacture, batteries powering BEV usually have a capacity between 20 kWh and 80 kWh, which ensure a range of up to 663 km [19,20]. Because the energy is supplied only by batteries, the BEV requires a larger battery pack size and capacity than other EVs.
HEVs are the most commonly used EV technology, combining an internal combustion engine (ICE) and an electric motor. In their powertrain, fossil fuel is the only external energy powering the ICE, and the energy for charging the battery is obtained from the braking process. The electric motor is used to power the vehicle for a short distance or to support the main engine (e.g., at a stoplight) [21]. Thus, HEVs require the lowest battery capacity, which is between 1.3 kWh and 1.6 kWh [19]. No charging plug for connection to the electricity grid exists in HEVs, which is usually considered to be a fuel-efficiency measure. Therefore, the battery capacity and size requirements are relatively low.
The powertrain of PHEVs is similar to the one of HEVs, and consists of an ICE and an electric engine; however, the main difference is that the PHEV can be plugged into an external source (e.g., EV charger) for charging. Furthermore, the HEV battery can also be charged throughout the regenerative braking processes. PHEVs have batteries with a larger capacity than HEVs, and, thus, they can extend the driving mileage when using electricity alone. Traditionally, this type of vehicle is powered by 50% gasoline and 50%
electricity. The size of the battery capacity for PHEV applications is usually between 4.5 kWh and 10 kWh [22].
There are many types of EVs in the market and each of them has individual specifics.
The most popular EVs in the market, classified according to their type, are summarized in Table 2 [19,23,24]. Furthermore, the global market share of HEV, PHEV, and BEV in 2020 is shown in Figure 2 [25].
Figure 1.The schematic representation of the three types of EVs; from left to right: Battery Electric Vehicles (BEV), Hybrid Electric Vehicles (HEV), and Plug-In Hybrid Electric Vehicles (PHEV).
BEVs are completely powered by electrical energy from batteries. The external energy is supplied only by plugging the BEV into the electricity grid for charging, and there is no on-board electricity generation. Depending on the EV manufacture, batteries powering BEV usually have a capacity between 20 kWh and 80 kWh, which ensure a range of up to 663 km [19,20]. Because the energy is supplied only by batteries, the BEV requires a larger battery pack size and capacity than other EVs.
HEVs are the most commonly used EV technology, combining an internal combustion engine (ICE) and an electric motor. In their powertrain, fossil fuel is the only external energy powering the ICE, and the energy for charging the battery is obtained from the braking process. The electric motor is used to power the vehicle for a short distance or to support the main engine (e.g., at a stoplight) [21]. Thus, HEVs require the lowest battery capacity, which is between 1.3 kWh and 1.6 kWh [19]. No charging plug for connection to the electricity grid exists in HEVs, which is usually considered to be a fuel-efficiency measure. Therefore, the battery capacity and size requirements are relatively low.
The powertrain of PHEVs is similar to the one of HEVs, and consists of an ICE and an electric engine; however, the main difference is that the PHEV can be plugged into an external source (e.g., EV charger) for charging. Furthermore, the HEV battery can also be charged throughout the regenerative braking processes. PHEVs have batteries with a larger capacity than HEVs, and, thus, they can extend the driving mileage when using electricity alone. Traditionally, this type of vehicle is powered by 50% gasoline and 50% electricity.
The size of the battery capacity for PHEV applications is usually between 4.5 kWh and 10 kWh [22].
There are many types of EVs in the market and each of them has individual specifics.
The most popular EVs in the market, classified according to their type, are summarized in Table2[19,23,24]. Furthermore, the global market share of HEV, PHEV, and BEV in 2020 is shown in Figure2[25].
Table 2.The comparison of the main characteristics of EVs in the market.
Type Manufacturer Battery-Only Range (km) Battery Capacity (kWh)
HEV
Toyota Prius IV
/ 1.3–1.6
BMW 225xe Audi A3 e-tron Toyota Prius III
PHEV
Chevy Volt 64 18.4
Toyota Prius XW30 21 5.2
Jaguar I-Pace 470 90.0
BEV
Tesla Model S 663 100.0
BMW i3 246 42.2
Nissan Leaf 364 62.0
Renault Zoe 395 52
Energies 2021, 14, x FOR PEER REVIEW 4 of 23
Table 2. The comparison of the main characteristics of EVs in the market.
Type Manufacturer Battery-Only Range (km) Battery Capacity (kWh)
HEV
Toyota Prius IV
/ 1.3–1.6
BMW 225xe Audi A3 e-tron Toyota Prius III PHEV
Chevy Volt 64 18.4
Toyota Prius XW30 21 5.2
Jaguar I-Pace 470 90.0
BEV
Tesla Model S 663 100.0
BMW i3 246 42.2
Nissan Leaf 364 62.0
Renault Zoe 395 52
Figure 2. The global market share of HEV, PHEV, and BEV in 2020 (based on [25]).
2.2. Lithium-Ion Chemistries in EV Applications
There is a growing demand for commercial LiBs with a high energy density for powering EVs. The energy density is mainly determined by the operating voltage of the active battery materials and their specific capacity. There are many types of LiBs with wide application prospects, such as spinel LMO/LNMO cathode [26], olivine LFP cathode [27,28], and layered NMC/NCA [29]. Among them, the group of Ni-rich layered oxides (LiNi1-xMxO2, M = Co, Mn, and Al) have both higher gravimetric and volumetric specific capacities than other intercalation-type cathode materials [30]. Moreover, the high working voltage of LiNi1-xMxO2 meets the withstand voltage of the current electrolytes.
Therefore, the NMC and NCA are widely used in current EV models [31]. More details of NMC and NCA-based batteries will be discussed in the following sections.
The performance of these Ni-rich cathode materials is greatly influenced by the properties of the used elements (i.e., Ni, Co, and Mn orAl). The capacity of the batteries is mainly provided by Ni; however, the use of Ni also leads to a poor cycle life and thermal stability. On the other hand, Mn and Al offer an improved cycle life and safety.
Furthermore, Co contributes to the electronic conductivity, ensuring a lower resistance.
The LiNixCoyMnzO2 (x = 1/3, 0.5, 0.6, 0.7, and 0.8) materials with high energy density, long cycle life, and excellent thermal stability have attracted much attention [32]. With the increasing of Ni content, the specific discharge capacity and total residual lithium also increase, but the corresponding capacity retention and safety characteristics gradually decrease [32,33]. However, this does not hinder its large-scale application in real life, due to the high reversible capacity, discharge working voltage, and relatively low costs for Ni- rich materials [34]. Among them, researchers replace the Mn element of LiNixCoyMnzO2 Figure 2.The global market share of HEV, PHEV, and BEV in 2020 (based on [25]).
Energies2021,14, 5220 4 of 22
2.2. Lithium-Ion Chemistries in EV Applications
There is a growing demand for commercial LiBs with a high energy density for powering EVs. The energy density is mainly determined by the operating voltage of the active battery materials and their specific capacity. There are many types of LiBs with wide application prospects, such as spinel LMO/LNMO cathode [26], olivine LFP cathode [27,28], and layered NMC/NCA [29]. Among them, the group of Ni-rich layered oxides (LiNi1−xMxO2, M = Co, Mn, and Al) have both higher gravimetric and volumetric specific capacities than other intercalation-type cathode materials [30]. Moreover, the high working voltage of LiNi1−xMxO2meets the withstand voltage of the current electrolytes.
Therefore, the NMC and NCA are widely used in current EV models [31]. More details of NMC and NCA-based batteries will be discussed in the following sections.
The performance of these Ni-rich cathode materials is greatly influenced by the properties of the used elements (i.e., Ni, Co, and Mn orAl). The capacity of the batteries is mainly provided by Ni; however, the use of Ni also leads to a poor cycle life and thermal stability. On the other hand, Mn and Al offer an improved cycle life and safety.
Furthermore, Co contributes to the electronic conductivity, ensuring a lower resistance.
The LiNixCoyMnzO2(x= 1/3, 0.5, 0.6, 0.7, and 0.8) materials with high energy density, long cycle life, and excellent thermal stability have attracted much attention [32]. With the increasing of Ni content, the specific discharge capacity and total residual lithium also increase, but the corresponding capacity retention and safety characteristics gradually decrease [32,33]. However, this does not hinder its large-scale application in real life, due to the high reversible capacity, discharge working voltage, and relatively low costs for Ni- rich materials [34]. Among them, researchers replace the Mn element of LiNixCoyMnzO2 with certain amounts of Al element [35]. The prepared LiNi0.8Co0.15Al0.05O2shows better thermal stability and cyclability than NMC cathodes [36].
The anode is used to store Li+ in the charging process. The traditional anode raw material is graphite (LiC6) with a capacity of 372 mAh/g. In addition, lithium titanate (LTO) with an operating voltage of around 1.55 V (vs. Li+/Li) has been considered for various vehicle applications. It enables the LTO to avoid self-discharge when working as an anode and thus enhances its safety and stability [37]. A brief comparative analysis of these chemistries is presented in Table3[30].
Table 3.A brief comparative analysis of common chemistries.
Electrode VAverage
(v)
VMax (v)
Specific Capacity (mAh g−1/mAh cm−3)
Gravimetric Energy (wh kg−1)
Cathode
LiNi0.8Co0.15Al0.05O2 3.7 4.6 220/979 758
LiNi1/3Co1/3Mn1/3O2 3.6 4.7 160/712 576
LiNi0.5Co0.2Mn0.3O2 3.6 4.7 170/757 612
LiNi0.6Co0.2Mn0.2O2 3.6 4.7 180/810 648
LiNi0.7Co0.15Mn0.15O2 3.6 4.7 190/855 684
LiNi0.8Co0.1Mn0.1O2 3.6 4.7 200/930 720
Anode graphite 0.6 3 372/735 190
LTO 1.55 2.5 175/607 263.5
3. Battery Operation in EV Application
The battery in EV applications is complex and diverse, as well as in the influence of various factors. It not only depends on driving habits, but also will be affected by environmental factors, such as the temperature, which varies from season-to-season and region-to-region. The load of the vehicle, driving frequency and mileage, charging habits and road conditions are different for different drivers, which puts the battery in a dif- ferent situation. Therefore, it is very difficult to standardize the battery operation in EV applications and even more difficult to estimate its state-of-health. However, the factors (e.g., driving habits, road conditions, load of the vehicles and so on) mainly affect battery
Energies2021,14, 5220 5 of 22
aging during driving. As a result, according to the battery working state, we can divide the battery usage status into three independent processes (charging process, driving, and standby) and study them separately, as illustrated in Figure3.
with certain amounts of Al element [35]. The prepared LiNi0.8Co0.15Al0.05O2 shows better thermal stability and cyclability than NMC cathodes [36].
The anode is used to store Li+ in the charging process. The traditional anode raw material is graphite (LiC6) with a capacity of 372 mAh/g. In addition, lithium titanate (LTO) with an operating voltage of around 1.55 V (vs. Li+/Li) has been considered for various vehicle applications. It enables the LTO to avoid self-discharge when working as an anode and thus enhances its safety and stability [37]. A brief comparative analysis of these chemistries is presented in Table 3 [30].
Table 3. A brief comparative analysis of common chemistries.
Electrode VAveragezqx (v)
VMaxzqx (v)
Specific Capacityzqx
(mAh g−1/mAh cm−3) Gravimetric Energy (wh kg−1)
Cathode
LiNi0.8Co0.15Al0.05O2 3.7 4.6 220/979 758
LiNi1/3Co1/3Mn1/3O2 3.6 4.7 160/712 576
LiNi0.5Co0.2Mn0.3O2 3.6 4.7 170/757 612
LiNi0.6Co0.2Mn0.2O2 3.6 4.7 180/810 648
LiNi0.7Co0.15Mn0.15O2 3.6 4.7 190/855 684
LiNi0.8Co0.1Mn0.1O2 3.6 4.7 200/930 720
Anode graphite 0.6 3 372/735 190
LTO 1.55 2.5 175/607 263.5
3. Battery Operation in EV Application
The battery in EV applications is complex and diverse, as well as in the influence of various factors. It not only depends on driving habits, but also will be affected by environmental factors, such as the temperature, which varies from season-to-season and region-to-region. The load of the vehicle, driving frequency and mileage, charging habits and road conditions are different for different drivers, which puts the battery in a different situation. Therefore, it is very difficult to standardize the battery operation in EV applications and even more difficult to estimate its state-of-health. However, the factors (e.g., driving habits, road conditions, load of the vehicles and so on) mainly affect battery aging during driving. As a result, according to the battery working state, we can divide the battery usage status into three independent processes (charging process, driving, and standby) and study them separately, as illustrated in Figure 3.
Figure 3. The battery operation modes in EV applications.
Figure 3.The battery operation modes in EV applications.
3.1. EV Battery Charging
The most used protocol for charging LiB is a constant current constant voltage (CC- CV), which consists of a constant current charging phase during which the battery voltage increases up to the cut-off value, followed by a constant voltage phase until the current falls to near-zero, as shown in Figure4a. The battery charging protocol is closely related to the daily usage of the EV and to the battery’s SOH, which will be reflected in the charging time. In order to reduce the charging time and negative effect on the battery from high currents, several alternative charging protocols are proposed in the literature, which are: Multistage Constant Current (MCC) charging, Pulse Charging (PC), Constant Power Constant Voltage (CP-CV), Variable Current Profile (VCP) and Constant Current Pulsed Charging (CC-PC) [38–41].
Energies 2021, 14, x FOR PEER REVIEW 6 of 23
3.1. EV Battery Charging
The most used protocol for charging LiB is a constant current constant voltage (CC- CV), which consists of a constant current charging phase during which the battery voltage increases up to the cut-off value, followed by a constant voltage phase until the current falls to near-zero, as shown in Figure 4a. The battery charging protocol is closely related to the daily usage of the EV and to the battery’s SOH, which will be reflected in the charging time. In order to reduce the charging time and negative effect on the battery from high currents, several alternative charging protocols are proposed in the literature, which are: Multistage Constant Current (MCC) charging, Pulse Charging (PC), Constant Power Constant Voltage (CP-CV), Variable Current Profile (VCP) and Constant Current Pulsed Charging (CC-PC) [38–41].
Figure 4. Schematic illustration of different charging protocols: (a) Constant Current Constant Voltage (CC-CV); (b) Mul- tistage Constant Current (MCC); (c) Pulse Charging (PC); (d) Constant Power Constant Voltage (CP-CV); (e) Variable Current Profile (VCP); (f) Constant Current Pulsed Charging (CC-PC).
To achieve fast charging and slow down the battery’s aging process, researchers proposed the Multistage Constant Current (MCC) protocol as one of the earliest charging types. This method sets different current levels during the charging process, as illustrated in Figure 4b, in order to minimize battery degradation. This is a very promising charging method, but the optimal value and duration of the current level need to be further researched [42]. To reduce the concentration polarization and mechanical stresses, charging is periodically interrupted by short rest periods (as presented in Figure 4c) or discharge pulses in the pulse charging protocols [41]. In [38], the authors found that the cycle life is similar to the battery when charged by CC-CV method; nevertheless, according to the results presented in [41], there is no overall agreement regarding the positive and/or negative effects on battery performance and lifetime of the pulsed charging methods. Besides, this charging method is complex, and also involves many factors to consider, such as pulsed current amplitude, its duty cycle, and pulse frequency, hindering its large-scale application [43]. The CP-CV protocol provides a low-current near the end of charging to alleviate metallic lithium plating at the anode (Figure 4d). However, as presented in [44], a fast capacity fading can be observed from a 0.5 C CP-CV charging, which is caused by high polarization in the initial CP-charging. Furthermore, some complex variable current profiles have been used for battery charging (Figure 4e). Sikha et al. designed, in [45], a varying current profile (VCP) protocol, which is less damaging Figure 4.Schematic illustration of different charging protocols: (a) Constant Current Constant Voltage (CC-CV); (b) Multi- stage Constant Current (MCC); (c) Pulse Charging (PC); (d) Constant Power Constant Voltage (CP-CV); (e) Variable Current Profile (VCP); (f) Constant Current Pulsed Charging (CC-PC).
Energies2021,14, 5220 6 of 22
To achieve fast charging and slow down the battery’s aging process, researchers pro- posed the Multistage Constant Current (MCC) protocol as one of the earliest charging types. This method sets different current levels during the charging process, as illustrated in Figure4b, in order to minimize battery degradation. This is a very promising charg- ing method, but the optimal value and duration of the current level need to be further researched [42]. To reduce the concentration polarization and mechanical stresses, charging is periodically interrupted by short rest periods (as presented in Figure4c) or discharge pulses in the pulse charging protocols [41]. In [38], the authors found that the cycle life is similar to the battery when charged by CC-CV method; nevertheless, according to the results presented in [41], there is no overall agreement regarding the positive and/or nega- tive effects on battery performance and lifetime of the pulsed charging methods. Besides, this charging method is complex, and also involves many factors to consider, such as pulsed current amplitude, its duty cycle, and pulse frequency, hindering its large-scale application [43]. The CP-CV protocol provides a low-current near the end of charging to alleviate metallic lithium plating at the anode (Figure4d). However, as presented in [44], a fast capacity fading can be observed from a 0.5 C CP-CV charging, which is caused by high polarization in the initial CP-charging. Furthermore, some complex variable current profiles have been used for battery charging (Figure4e). Sikha et al. designed, in [45], a varying current profile (VCP) protocol, which is less damaging than the pure CV charging in the early cycles. However, this profile, too, resulted in a significant capacity degradation compared to that of CC-CV with the same mean current. Another battery charging protocol is the CC-PC charging protocol which combines a constant current and pulsed charging (Figure4f). The application cost of the charging protocol is relatively low, avoiding voltage control and variable charging currents [40]. But, it can deteriorate the cycle life due to pulses leading to an exceedance voltage in the high voltage period [38].
At present, the standardized CC-CV protocol is widely applied in real-life applications (e.g., EVs), while the other presented charging protocols, developed based on an array of physical motivations which include plenty of advantages in some certain conditions (such as fast-charging), are mainly applied at the research stage or even at a reduced-scale in real life. Currently, the EVs are mainly charged by the CC-CV protocol. In the aging analysis section (Sections4and5), we mainly discuss the aging mechanisms which are caused by the CC-CV charging protocols.
3.2. EV Driving Operation 3.2.1. Real-Life Scenarios
The driving profiles of EVs in real life are diverse. They mainly depend on the vehicle design, the behavior of the driver, and the external environment, resulting in difference in service life and health status. Generally, the vehicle design considers multiple factors in order to meet the needs of different consumers, including the overall dimensions, passenger capacity, tire type and shape, specific front area, and body type [46]. This will make the EVs operate differently in different situations, and lead to different energy demands. In addition, the battery pack sizing and output power are also different. Driver behavior is also related to different traffic conditions. Furthermore, the external environment conditions (e.g., mainly ambient temperature, but also precipitation, wind, etc.) also have an important effect on the battery loading and, subsequently, on the range of the EV. All of these factors make EV driving very diverse. Therefore, it is difficult to study, analyze, and propose a unified set of aging mechanisms for batteries operated in EVs using real-life data. There is limited available literature regarding battery usage harvested from real-life operations. For example, Jafari et al. analyzed, in [47], the real-world daily driving undertaken by a fleet of connected vehicles. The data set contains the records of the connected vehicles volunteered by drivers using their vehicles in the US. The data from 50 driving cycles of 50 drivers were used to evaluate EV driving during an entire year.
In order to simulate the driving conditions of the EVs and evaluate the degradation behavior of the batteries, researchers developed a series of driving cycle profiles based on
Energies2021,14, 5220 7 of 22
extensive real-life data, which fully considers various factors (e.g., driving habits, road conditions, load of the vehicles, vehicles type, weather, and so on). These standardized EV driving profiles are very convenient for laboratory battery testing and performance- degradation evaluation.
3.2.2. Driving Cycle Profiles
Many countries and organizations create their standard driving cycle based on their roads and environment, in order to assess the performance of EVs and their batteries [44].
The driving cycle consists of a series of repetitive sequences of vehicle operating modes that represent the driving models, as shown in Figure5.
EURO 6, which is the most recent emission standard for passenger cars imposed by the European Union Commission, the light-duty vehicles are tested with an updated standard driving cycle, the World Harmonized Light Vehicles Test Cycle (WLTC), which provides more dynamic driving behaviors [48]. WLTC was generated to define a globally harmo- nized standard for determining the levels of pollutants, energy consumption, and the elec- tric range for light-duty driving, from approximately 765,000 km of data gathered in five different regions: EU, USA, India, Korea, and Japan. The “real world” driving data con- tains different road characteristics (urban, rural, and motorway) and driving conditions (peak, off-peak, weekend), covering a wide range of vehicle categories, various engines and manufacturers. This is similar to the different kinds of roads that are commonly trav- elled on, containing a low speed (L/ < 60 km/h), a medium (M/ < 80 km/h), a high speed (H/ < 110 km/h) and extra-high speed (Ex-H/ > 110 km/h), corresponding to the urban, rural and motorway classification, as shown in Figure 5c. Nowadays, more and more manufactures use WLTC to evaluate the performance of PHEV, HEV and BEV. In below table, a comparison of the aforementioned driving cycles is presented.
Figure 5. The driving cycle profiles of (a) UDDS, (b) NEDC and (c) WLTC for a class 3 vehicle.
To prove the validity of the WLTC profile for EVs studies, Ben-Marzouk et al. [49]
collected the real-life EV data for two years on ten EVs, and built synthetic profiles of the real application M1 and real application M2, as shown in Figure 6. These profiles are compared with the WLTC profile and they show great similarity in terms of aging, especially for the real application M1. These results validate the use of the WLTC profile in aging tests and aging studies of EV batteries.
Figure 5.The driving cycle profiles of (a) UDDS, (b) NEDC and (c) WLTC for a class 3 vehicle.
The EPA Urban Dynamometer Driving Schedule (UDDS), presented in Figure5a, is commonly called the “LA4” or “the city test”; the UDDS represents city driving conditions and has been used as the standard driving cycle in United States since 1972 for light-duty vehicle testing. The New European Driving Cycle (NEDC) is also an early driving cycle that was introduced in 1996 and divided into two parts, as shown in Figure5b. The first part simulates urban driving conditions which are repeated four times and the second part simulates extra-urban driving conditions with a high speed. With the regulation of EURO 6, which is the most recent emission standard for passenger cars imposed by the European Union Commission, the light-duty vehicles are tested with an updated standard driving cycle, the World Harmonized Light Vehicles Test Cycle (WLTC), which provides more dynamic driving behaviors [48]. WLTC was generated to define a globally harmonized standard for determining the levels of pollutants, energy consumption, and the electric range for light-duty driving, from approximately 765,000 km of data gathered in five different regions: EU, USA, India, Korea, and Japan. The “real world” driving data contains different road characteristics (urban, rural, and motorway) and driving conditions (peak, off-peak, weekend), covering a wide range of vehicle categories, various engines and manufacturers. This is similar to the different kinds of roads that are commonly travelled on, containing a low speed (L/ < 60 km/h), a medium (M/ < 80 km/h), a high speed (H/ < 110 km/h) and extra-high speed (Ex-H/ > 110 km/h), corresponding to the urban, rural and motorway classification, as shown in Figure5c. Nowadays, more and more manufactures use WLTC to evaluate the performance of PHEV, HEV and BEV. In below table, a comparison of the aforementioned driving cycles is presented.
To prove the validity of the WLTC profile for EVs studies, Ben-Marzouk et al. [49]
collected the real-life EV data for two years on ten EVs, and built synthetic profiles of the real application M1 and real application M2, as shown in Figure6. These profiles
Energies2021,14, 5220 8 of 22
are compared with the WLTC profile and they show great similarity in terms of aging, especially for the real application M1. These results validate the use of the WLTC profile in aging tests and aging studies of EV batteries.
Energies 2021, 14, x FOR PEER REVIEW 8 of 23
EURO 6, which is the most recent emission standard for passenger cars imposed by the European Union Commission, the light-duty vehicles are tested with an updated standard driving cycle, the World Harmonized Light Vehicles Test Cycle (WLTC), which provides more dynamic driving behaviors [48]. WLTC was generated to define a globally harmo- nized standard for determining the levels of pollutants, energy consumption, and the elec- tric range for light-duty driving, from approximately 765,000 km of data gathered in five different regions: EU, USA, India, Korea, and Japan. The “real world” driving data con- tains different road characteristics (urban, rural, and motorway) and driving conditions (peak, off-peak, weekend), covering a wide range of vehicle categories, various engines and manufacturers. This is similar to the different kinds of roads that are commonly trav- elled on, containing a low speed (L/ < 60 km/h), a medium (M/ < 80 km/h), a high speed (H/ < 110 km/h) and extra-high speed (Ex-H/ > 110 km/h), corresponding to the urban, rural and motorway classification, as shown in Figure 5c. Nowadays, more and more manufactures use WLTC to evaluate the performance of PHEV, HEV and BEV. In below table, a comparison of the aforementioned driving cycles is presented.
Figure 5. The driving cycle profiles of (a) UDDS, (b) NEDC and (c) WLTC for a class 3 vehicle.
To prove the validity of the WLTC profile for EVs studies, Ben-Marzouk et al. [49]
collected the real-life EV data for two years on ten EVs, and built synthetic profiles of the real application M1 and real application M2, as shown in Figure 6. These profiles are compared with the WLTC profile and they show great similarity in terms of aging, especially for the real application M1. These results validate the use of the WLTC profile in aging tests and aging studies of EV batteries.
Figure 6. Evolution of the capacity of the batteries according to Ah exchanged for each aging profile [49].
3.3. EV Standby Operation
In EV applications, the standby state accounts for a large proportion of the entire battery service life. As presented in [50], the standby time can reach 90% of the total vehicle operation. This means that calendar aging cannot be ignored when assessing battery degradation and lifetime in such applications. In [51], Swierczynski et al. showed that more than 75% of capacity fade was caused by calendar aging during the EV operation.
Thus, it is particularly important to study the calendar aging process and related factors, such as the state-of-charge (SOC) and temperature. In normal conditions, the calendar aging process is slow, and it takes several years or even tens of years for a battery to end its service life by calendaring only. In actual research, elevated temperature and/or high SOC levels are often used to accelerate battery aging.
4. Aging in EV Application 4.1. Aging in Charging
The LiBs aging in the charging process is related to many factors, such as the cut-off voltage, current rate, and temperature. Using a high cut-off voltage, more capacity can be charged and a high charging current can significantly shorten the charging period.
However, in these cases, the degradation process of LiBs will be greatly accelerated. The most common factors affecting the charging process and their variations are shown in Figure7.
4.1.1. Impact of Charging Voltage
Slight overcharging or reduction in the cut-off voltage have an obvious impact on the degradation of LiBs. In [52] the author accelerated the battery aging by 300 mV overcharging. The schematic diagram is shown in Figure8a. They designed a cycle aging experiment for 18650-type batteries (i.e., 2 Ah, NMC/graphite) and charged one of them in the voltage range of 2.75–4.5 V, while another battery in was charged in the nominal range of 2.75–4.2 V. The battery charging/discharging profiles are shown in Figure8b. As expected, the overcharged battery has an 18% higher capacity than the battery charged in the nominal conditions; however, for the overcharged batteries (three overcharged cells tested at the same condition) the capacity fade is pronounced, as shown in Figure8c. The SOH drops to 80% after only approximately 40 cycles, which is much quickly than for the battery aged under normal voltage (which needed over 100 cycles to reach the same SOH).
Thus, overcharging the battery highly accelerates the degradation process.
Energies2021,14, 5220 9 of 22 Figure 6. Evolution of the capacity of the batteries according to Ah exchanged for each aging profile [49].
3.3. EV Standby Operation
In EV applications, the standby state accounts for a large proportion of the entire battery service life. As presented in [50], the standby time can reach 90% of the total vehicle operation. This means that calendar aging cannot be ignored when assessing battery degradation and lifetime in such applications. In [51], Swierczynski et al. showed that more than 75% of capacity fade was caused by calendar aging during the EV operation. Thus, it is particularly important to study the calendar aging process and related factors, such as the state-of-charge (SOC) and temperature. In normal conditions, the calendar aging process is slow, and it takes several years or even tens of years for a battery to end its service life by calendaring only. In actual research, elevated temperature and/or high SOC levels are often used to accelerate battery aging.
4. Aging in EV Application 4.1. Aging in Charging
The LiBs aging in the charging process is related to many factors, such as the cut-off voltage, current rate, and temperature. Using a high cut-off voltage, more capacity can be charged and a high charging current can significantly shorten the charging period.
However, in these cases, the degradation process of LiBs will be greatly accelerated. The most common factors affecting the charging process and their variations are shown in Figure 7.
Figure 7. The influence on the battery aging from the factors of the charging process.
4.1.1. Impact of Charging Voltage
Slight overcharging or reduction in the cut-off voltage have an obvious impact on the degradation of LiBs. In [52] the author accelerated the battery aging by 300 mV overcharging. The schematic diagram is shown in Figure 8a. They designed a cycle aging experiment for 18650-type batteries (i.e., 2 Ah, NMC/graphite) and charged one of them in the voltage range of 2.75–4.5 V, while another battery in was charged in the nominal range of 2.75–4.2 V. The battery charging/discharging profiles are shown in Figure 8b. As Figure 7.The influence on the battery aging from the factors of the charging process.
Energies 2021, 14, x FOR PEER REVIEW 10 of 23
expected, the overcharged battery has an 18% higher capacity than the battery charged in the nominal conditions; however, for the overcharged batteries (three overcharged cells tested at the same condition) the capacity fade is pronounced, as shown in Figure 8c. The SOH drops to 80% after only approximately 40 cycles, which is much quickly than for the battery aged under normal voltage (which needed over 100 cycles to reach the same SOH).
Thus, overcharging the battery highly accelerates the degradation process.
In contrast, in [53], Rathieu et al. charged two different types of batteries with a slight reduction of 100 mV in the cut-off voltage, as shown in Figure 8d. For battery A (i.e., 3.0 Ah, NMC 811/G-SiO), the reduction of the cut-off voltage led to both a lower charging time and a lower degradation (Figure 8e); however, this results in a reduction of the charged capacity (i.e., only 83%). For battery B (i.e., 2.5 Ah, NCA/G), there is no significant reduction in the charging time or in the degradation (Figure 8f), but a lower capacity is achieved (i.e., 89%).
Furthermore, in [54], the author analyzed the contribution of the constant voltage process to the battery charging, and charged 18650-type batteries (2.5 Ah, NMC/Graphite, and NCA/Graphite) by the CC (2.5 A to 4.2 V) (Figure 8g) and CC-CV protocols (2.5 A to 4.2 V and 0.1 A cut-off current) at room temperature, respectively. The battery charged with the CC-CV protocol lost 20% of its capacity after 500 cycles, while the battery charged only with the CC protocol reached the same degradation level after 600 cycles, as shown in Figure 8h. From Figure 8i, it can be seen that the batteries contribute a similar total capacity throughput charged by CC-CV and CC protocols. However, using the CC-CV protocol, more capacity (about 20%) can be charged in comparison to the CC protocol.
Figure 8. The effect of the charging voltage on the battery capacity fade. (a) the schematic diagram of slight overcharging;
(b) charging/discharging profiles for standard and overcharge cells, (c) capacity fading of three LiBs under the same over- charging cycling condition (2.75–4.5 V) [52]. (d) the schematic diagram of reduction in cut-off voltage; (e) the capacity fade curves for cell A, (f) the capacity fade curves for cell B [53]. (g) the schematic diagram of CC charging protocol; (h) capacity degradation curves based on cycles; (i) capacity degradation curves based on throughput [54].
Figure 8.The effect of the charging voltage on the battery capacity fade. (a) the schematic diagram of slight overcharging;
(b) charging/discharging profiles for standard and overcharge cells, (c) capacity fading of three LiBs under the same overcharging cycling condition (2.75–4.5 V) [52]. (d) the schematic diagram of reduction in cut-off voltage; (e) the capacity fade curves for cell A, (f) the capacity fade curves for cell B [53]. (g) the schematic diagram of CC charging protocol;
(h) capacity degradation curves based on cycles; (i) capacity degradation curves based on throughput [54].
Energies2021,14, 5220 10 of 22
In contrast, in [53], Rathieu et al. charged two different types of batteries with a slight reduction of 100 mV in the cut-off voltage, as shown in Figure8d. For battery A (i.e., 3.0 Ah, NMC 811/G-SiO), the reduction of the cut-off voltage led to both a lower charging time and a lower degradation (Figure8e); however, this results in a reduction of the charged capacity (i.e., only 83%). For battery B (i.e., 2.5 Ah, NCA/G), there is no significant reduction in the charging time or in the degradation (Figure8f), but a lower capacity is achieved (i.e., 89%).
Furthermore, in [54], the author analyzed the contribution of the constant voltage process to the battery charging, and charged 18650-type batteries (2.5 Ah, NMC/Graphite, and NCA/Graphite) by the CC (2.5 A to 4.2 V) (Figure8g) and CC-CV protocols (2.5 A to 4.2 V and 0.1 A cut-off current) at room temperature, respectively. The battery charged with the CC-CV protocol lost 20% of its capacity after 500 cycles, while the battery charged only with the CC protocol reached the same degradation level after 600 cycles, as shown in Figure8h. From Figure8i, it can be seen that the batteries contribute a similar total capacity throughput charged by CC-CV and CC protocols. However, using the CC-CV protocol, more capacity (about 20%) can be charged in comparison to the CC protocol.
It can be concluded that slight overcharging will increase the capacity but will obvi- ously accelerate the battery aging. For the reduction in cut-off voltage and the removing of the CV charging, there is no significant reduction in the charging time and in the degradation, but a lower capacity is achieved.
4.1.2. Impact of the Charging Current
In charging, the currentIchis an important parameter for charging speed. Spingler et al. [55] charged batteries (i.e., NMC/graphite, 3.3 Ah) with a constant current between 1.0 C and 2.0 C. The capacity loss and both average and maximum local irreversible expansion per cycle are shown in Figure9a. With the increase of the C-rate, the battery capacity fade becomes more and more obvious. Furthermore, the irreversible expansion and capacity loss are correlated with a bivariate correlation of 0.996. This means that a higher charging rate will lead to a larger irreversible expansion, corresponding to a growing tendency for capacity loss. Especially for the 2 C charging, the fade of capacity is about 30 mAh per cycle.
Energies 2021, 14, x FOR PEER REVIEW 11 of 23
It can be concluded that slight overcharging will increase the capacity but will obvi- ously accelerate the battery aging. For the reduction in cut-off voltage and the removing of the CV charging, there is no significant reduction in the charging time and in the deg- radation, but a lower capacity is achieved.
4.1.2. Impact of the Charging Current
In charging, the current Ich is an important parameter for charging speed. Spingler et al. [55] charged batteries (i.e., NMC/graphite, 3.3 Ah) with a constant current between 1.0 C and 2.0 C. The capacity loss and both average and maximum local irreversible expansion per cycle are shown in Figure 9a. With the increase of the C-rate, the battery capacity fade becomes more and more obvious. Furthermore, the irreversible expansion and capacity loss are correlated with a bivariate correlation of 0.996. This means that a higher charging rate will lead to a larger irreversible expansion, corresponding to a growing tendency for capacity loss. Especially for the 2 C charging, the fade of capacity is about 30 mAh per cycle.
Figure 9. (a) Capacity loss and irreversible expansion per cycle as a function of C-rate in CC-CV charging [55]; (b,c) cycle life for different combinations of moderate (3 A) and high (5 A) charging current Ich and discharging current Idis [38]; (d,e) impact of charge current Ich on cycle life at 25 °C for two types of batteries [53].
Keil et al. analyzed the impact of high charging currents on the cycle life of two different types of batteries: Model A (i.e., (LMO+NMC)/graphite) and Model B (i.e., (NMC+LCO)/graphite), as shown in Figure 9b,c, respectively [38]. For the Model A battery cell (Figure 9b), the increase in the charging C-rate (from 1 C to 5 C) increases the battery degradation (i.e., capacity fade). On the other hand, for the Model B battery cells, (Figure 9c) the battery degradation is not influenced by the charging C-rate, which may be related to the composition of the cathode.
Furthermore, in [53], the authors compared the influence of three different charging currents (i.e., 3 A, 4 A, and 5 A) on two LiBs, cell A (i.e., NMC/graphite) and cell B (i.e., NCA/graphite), considering two cells for each aging condition. The capacity degradation tendency, which is presented in Figure 9d (cell A) and Figure 9e (cell B), is similar for the two investigated chemistries; in the beginning, there is a high capacity loss, followed by a slower capacity fade in the middle of the life cycle, and ending with a sudden capacity degradation period. For cell A, the capacity fade behavior of the battery cells charged with 5 A and 4 A is similar, and a faster degradation is observed than for the cells charged with Figure 9.(a) Capacity loss and irreversible expansion per cycle as a function of C-rate in CC-CV charging [55]; (b,c) cycle life for different combinations of moderate (3 A) and high (5 A) charging currentIchand discharging currentIdis[38];
(d,e) impact of charge currentIchon cycle life at 25◦C for two types of batteries [53].
Keil et al. analyzed the impact of high charging currents on the cycle life of two different types of batteries: Model A (i.e., (LMO+NMC)/graphite) and Model B (i.e., (NMC+LCO)/graphite), as shown in Figure9b,c, respectively [38]. For the Model A battery cell (Figure9b), the increase in the charging C-rate (from 1 C to 5 C) increases the battery degradation (i.e., capacity fade). On the other hand, for the Model B battery cells, (Figure9c) the battery degradation is not influenced by the charging C-rate, which may be related to the composition of the cathode.
Furthermore, in [53], the authors compared the influence of three different charging currents (i.e., 3 A, 4 A, and 5 A) on two LiBs, cell A (i.e., NMC/graphite) and cell B (i.e., NCA/graphite), considering two cells for each aging condition. The capacity degradation tendency, which is presented in Figure9d (cell A) and Figure9e (cell B), is similar for the two investigated chemistries; in the beginning, there is a high capacity loss, followed by a slower capacity fade in the middle of the life cycle, and ending with a sudden capacity degradation period. For cell A, the capacity fade behavior of the battery cells charged with 5 A and 4 A is similar, and a faster degradation is observed than for the cells charged with 3 A (especially after 20% capacity fade). On the other hand, it can be observed that the degradation paths are similar for cell B under all the currents from 3 A to 5 A (Figure9e).
Based on the above studies, it can be concluded that, in most cases, higher charging current rates result in the faster degradation of LiBs; however, both the degradation trend and the influence of the charging rate change from chemistry-to-chemistry.
4.1.3. Impact of Charging Temperature
At low-temperatures, LiBs are characterized. In [56], the authors assessed the battery (i.e., NCA/graphite, 2.9 Ah) charging in different low temperatures (i.e.,−5◦C,−10◦C,
−15◦C and−20◦C). After charging to full state, they set a 3 h rest at room temperature for the batteries, and then discharged all the batteries at 25◦C. At low temperatures, not only is the charging efficiency lowered, but the energy that can be charged is correspondingly reduced. Moreover, the authors found that the capacity obviously declines, with the reduction in the ambient charging temperature. In the harsh conditions of−20◦C, the battery capacity drops to 72% after only eight cycles. In contrast, at higher temperatures (>+40◦C), the charging capacity increases and the internal resistance decreases further, compared with charging at room temperature [57].
To conclude, the aging of the battery during the charging process is affected by the cut-off voltage, current and temperature. High cut-off voltage, high current and extreme temperatures (both low and high temperatures) will accelerate the battery aging.
4.2. Aging in Driving
The battery degradation in EV driving is sensitive to a variety of factors from the behavior of the user to the environmental conditions. The mileage of the EV’s daily use is related to the battery depth of discharge (DOD); driving speed and acceleration are related to the battery discharge current; and the main environmental factor that influences the battery aging is the temperature. These factors and the corresponding influences on the battery are summarized in Figure10.
Performing laboratory accelerated aging test is an effective method to analyze degra- dation in EV batteries. In [58], Stroe et al. carried out a daily aging profile (e.g., WLTC), which consisted of 22 h cycling and 2 h stand-by, as presented in Figure11a, and analyz- ing and assessing the aging of NMC-based battery cells. Furthermore, the temperature changed monthly in accordance with the climate of Seville, Spain. After eleven months of accelerated aging, the two tested cells lost approximately 10% of their initial capacity (Figure11b). A slow down tendency of the capacity fade appeared as the aging of the bat- tery evolved, and the authors attributed this to the irreversible capacity loss caused by the formation and growth of the solid electrolyte interface (SEI) layer in the initial stages. The relationship between the monthly capacity fade and the temperature is shown in Figure11c.
The highest capacity fade occurs in the first month, approximately 3%, corresponding to
Energies2021,14, 5220 12 of 22
a cycling temperature of 36◦C. They also found that the degradation is minimal when the battery is cycled at 26◦C (approximately room temperature). In addition, the increase of internal resistance accelerates as the aging process evolves during the eleven months of testing (Figure11d). In [59], the author obtained a similar conclusion that the capacity of pouch cell (e.g., NMC/graphite, 20 Ah) will fade more quickly in high-temperature driving. They show that after 2184 cycles, the battery lost 19.2% capacity when aged using the WLTC at 45◦C. On the contrary, when the battery was aged at a temperature of 10◦C, almost no capacity degradation was observed after 1428 cycles.
Energies 2021, 14, x FOR PEER REVIEW 12 of 23
3 A (especially after 20% capacity fade). On the other hand, it can be observed that the degradation paths are similar for cell B under all the currents from 3 A to 5 A (Figure 9e).
Based on the above studies, it can be concluded that, in most cases, higher charging current rates result in the faster degradation of LiBs; however, both the degradation trend and the influence of the charging rate change from chemistry-to-chemistry.
4.1.3. Impact of Charging Temperature
At low-temperatures, LiBs are characterized. In [56], the authors assessed the battery (i.e., NCA/graphite, 2.9 Ah) charging in different low temperatures (i.e., −5 °C, −10 °C, −15
°C and −20 °C). After charging to full state, they set a 3 h rest at room temperature for the batteries, and then discharged all the batteries at 25 °C. At low temperatures, not only is the charging efficiency lowered, but the energy that can be charged is correspondingly reduced. Moreover, the authors found that the capacity obviously declines, with the reduction in the ambient charging temperature. In the harsh conditions of −20 °C, the battery capacity drops to 72% after only eight cycles. In contrast, at higher temperatures (>+40 °C), the charging capacity increases and the internal resistance decreases further, compared with charging at room temperature [57].
To conclude, the aging of the battery during the charging process is affected by the cut- off voltage, current and temperature. High cut-off voltage, high current and extreme temperatures (both low and high temperatures) will accelerate the battery aging.
4.2. Aging in Driving
The battery degradation in EV driving is sensitive to a variety of factors from the behavior of the user to the environmental conditions. The mileage of the EV’s daily use is related to the battery depth of discharge (DOD); driving speed and acceleration are related to the battery discharge current; and the main environmental factor that influences the battery aging is the temperature. These factors and the corresponding influences on the battery are summarized in Figure 10.
Figure 10. The influence in battery aging from common factors in driving (WLTC) test.
Performing laboratory accelerated aging test is an effective method to analyze degradation in EV batteries. In [58], Stroe et al. carried out a daily aging profile (e.g., WLTC), which consisted of 22 h cycling and 2 h stand-by, as presented in Figure 11a, and analyzing and assessing the aging of NMC-based battery cells. Furthermore, the temperature changed monthly in accordance with the climate of Seville, Spain. After eleven months of accelerated aging, the two tested cells lost approximately 10% of their initial capacity (Figure 11b). A slow down tendency of the capacity fade appeared as the aging of the battery evolved, and the authors attributed this to the irreversible capacity loss caused by the formation and growth of the solid electrolyte interface (SEI) layer in the initial stages. The relationship between the monthly capacity fade and the temperature is shown in Figure 11c. The highest capacity fade occurs in the first month, approximately
Figure 10.The influence in battery aging from common factors in driving (WLTC) test.
Energies 2021, 14, x FOR PEER REVIEW 13 of 23
3%, corresponding to a cycling temperature of 36 °C. They also found that the degradation is minimal when the battery is cycled at 26 °C (approximately room temperature). In addition, the increase of internal resistance accelerates as the aging process evolves during the eleven months of testing (Figure 11d). In [59], the author obtained a similar conclusion that the capacity of pouch cell (e.g., NMC/graphite, 20 Ah) will fade more quickly in high- temperature driving. They show that after 2184 cycles, the battery lost 19.2% capacity when aged using the WLTC at 45 °C. On the contrary, when the battery was aged at a temperature of 10 °C, almost no capacity degradation was observed after 1428 cycles.
In [58], it was predicted that the battery will reach 20% capacity fade under the considered aging profile (based on the WLTC driving cycle) after approximately 5.8 years.
To further accelerate battery aging, Simolka et al. selected the last part of the WLTC driving cycle (called “extra high”), corresponding to high currents, and named it as “AP1”
[60]. Based on AP1, the authors then doubled the current load in order to create a new profile: “AP2”. The capacity fade of the LFP/G cells was tested based on AP1 and AP2 profiles; at the same time, two DODs (i.e., 50% and 100%) were performed. These methods are very suitable for studying the real-life applications of batteries in EVs. However, there are few studies about NMC and NCA battery discharge based on these methods, currently.
Figure 11. (a) One−day load current profile for accelerated aging based on WLTC driving cycle; (b)
capacity fade of the tested NMC-based cells during eleven months of accelerated aging tests; (c) correlation between the monthly measured battery capacity fade and the considered temperature;
(d) measured internal resistance increase at different SOC [58].
Overdischarging also leads to the degradation of the battery. In [61], the author found that overdischarging deteriorates the electrode materials. On the one hand, as the discharging cut-off voltage decreases, the surface temperature of the battery increases significantly, leading to transition metal dissolution at cathode; on the other hand, overdischarging causes irreversible structural transformation of cathode and anode, resulting in a decrease in capacity. As presented by Lai et al. in [62], the critical over- discharging range for the DOD is from 115% to 120%. When over-discharging exceeds this range (i.e., more than 20% overdischarging), the rate of capacity degradation is greatly accelerated. If the open-circuit voltage of the battery does not recover to values higher than 2 V during the rest process, this may lead to an irreversible internal short circuit.
Figure 11. (a) One−day load current profile for accelerated aging based on WLTC driving cycle;
(b) capacity fade of the tested NMC-based cells during eleven months of accelerated aging tests;
(c) correlation between the monthly measured battery capacity fade and the considered temperature;
(d) measured internal resistance increase at different SOC [58].
In [58], it was predicted that the battery will reach 20% capacity fade under the considered aging profile (based on the WLTC driving cycle) after approximately 5.8 years.
To further accelerate battery aging, Simolka et al. selected the last part of the WLTC driving cycle (called “extra high”), corresponding to high currents, and named it as “AP1” [60].
Based on AP1, the authors then doubled the current load in order to create a new profile:
“AP2”. The capacity fade of the LFP/G cells was tested based on AP1 and AP2 profiles; at the same time, two DODs (i.e., 50% and 100%) were performed. These methods are very suitable for studying the real-life applications of batteries in EVs. However, there are few studies about NMC and NCA battery discharge based on these methods, currently.
Overdischarging also leads to the degradation of the battery. In [61], the author found that overdischarging deteriorates the electrode materials. On the one hand, as the discharg- ing cut-off voltage decreases, the surface temperature of the battery increases significantly, leading to transition metal dissolution at cathode; on the other hand, overdischarging causes irreversible structural transformation of cathode and anode, resulting in a decrease in capacity. As presented by Lai et al. in [62], the critical over-discharging range for the DOD is from 115% to 120%. When over-discharging exceeds this range (i.e., more than 20%
overdischarging), the rate of capacity degradation is greatly accelerated. If the open-circuit voltage of the battery does not recover to values higher than 2 V during the rest process, this may lead to an irreversible internal short circuit. Even worse, an explosion caused by overdischarge when the external temperature is high has been reported [63].
It can be concluded that high temperatures will accelerate battery degradation in real-life driving. Furthermore, overdischarging will also lead to battery degradation.
4.3. Aging in Standby
The standby state takes up a considerable amount of time in EVs’ real-life operation, and the contribution to total aging is significant. The pace of the degradation process during standby, known as calendar aging, varies depending on the SOC and temperature.
The effect of the SOC and temperature on the capacity fade of NMC-based LiBs is exemplified in Figure12a [64]. It is evident that the battery capacity fade is more notable at high temperatures than at a moderate temperature. In addition, high SOCs will also accelerate capacity fade due to the high voltage. As expected, the highest capacity degradation occurs in the toughest conditions, with a high storage SOC (80%), and the highest temperatures (45◦C) among these cases. The calendar aging situation is presented in Figure12b. In [65], the authors varied the storage SOC from 5% to 95%, and obtained a similar conclusion that a higher SOC (i.e., over 70%) will accelerate the battery (e.g., LTO/NMC) aging significantly during 300 days of calendar aging. However, there is no evident capacity degradation in cells with a SOC below 70%, even at 60 ◦C. This behavior occurs because the LTO/NMC cells are more stable than other cells, even at high temperatures. However, as reported in [66], when the cells (i.e., LMO+NMC/graphite) were stored at a middle SOC (i.e., 50%), the degradation is faster than for the cells which were stored at extreme SOCs (i.e., 10% and 90%). This is different from what was presented in previous literature that showed an increase in degradation by increasing the SOC at 25◦C. However, it is obvious from most of the available literature that storage/idling the LiBs at low SOCs and in low temperatures will result in slower capacity fade and, subsequently, a long lifetime. Some of the recent work investigating the calendar aging of NMC- and NCA-based LiBs is summarized in Table4.
